Dhaka University Journal of Applied Science and Engineering

Characterization of the CZTS (Cu2ZnSnS4) thin film for solar cell absorber layer synthesized from the nitrate-based sol-gel precursor solution

M Rahman — 2023-04-04

Manufacture of eco-friendly CZTS (Cu2ZnSnS4) thin film absorption layer of pure keserite structure for solar cell is far-reaching. We have manufactured the CZTS thin films by sol-gel dip-coating means from the nitrate-based chemicals. After hardening at 550°C in vacuum condition, we characterize the films by UV–vis spectroscopy, X-ray diffraction (XRD), Scanning Electron Microscope (SEM), and energy dispersive X-ray spectroscopy (EDS) methods. The CZTS films provided high optical absorption coefficient (2.92x104cm-1) and average band gap energy (1.58 eV). X-ray diffraction analysis proved the kesterite structure of films. The surface morphology analysis proved the deposition of crammed, condensed and granulated CZTS films. The thin films have intermittent disposal of agglomerated particles with clear-cut edges. The energy dispersive X-ray spectroscopy analysis conferred stoichiometric ratio as Cu: Zn: Sn: S= 2.1: 1.3: 1: 5.2.

DUJASE Vol. 7 (2) 1-7, 2022 (July)

Context-based Bengali Next Word Prediction: A Comparative Study of Different Embedding Methods

Mahir Mahbub — 2023-04-04

Next word prediction is a helpful feature for various typing subsystems. It is also convenient to have suggestions while typing to speed up the writing of digital documents. Therefore, researchers over time have been trying to enhance the capability of such a prediction system. Knowledge regarding the inner meaning of the words along with the contextual understanding of the sequence can be helpful in enhancing the next word prediction capability. Theoretically, these reasonings seem to be very promising. With the advancement of Natural Language Processing (NLP), these reasonings are found to be applicable in real scenarios. NLP techniques like Word embedding and sequential contextual modeling can help us to gain insight into these points. Word embedding can capture various relations among the words and explain their inner knowledge. On the other hand, sequence modeling can capture contextual information. In this paper, we figure out which embedding method works better for Bengali next word prediction. The embeddings we have compared are word2vec skip-gram, word2vec CBOW, fastText skip-gram and fastText CBOW. We have applied them in a deep learning sequential model based on LSTM which was trained on a large corpus of Bengali texts. The results reveal some useful insights about the contextual and sequential information gathering that will help to implement a context-based Bengali next word prediction system.

DUJASE Vol. 7 (2) 8-15, 2022 (July)

Design and Development of a Low-Cost 2.4 GHz GFSK Modulation based Walkie-Talkie

Galib Hashmi — 2023-04-04

A walkie-talkie device can be quite useful in the location where there is no cellphone coverage. Even in the cellphone coverage area it provides communication at very low cost. Currently the law enforcement and military personnel uses frequency modulation (FM) based walkie-talkie systems. The main goal of this research is to design and develop a lowcost walkie-talkie device that can be used by anyone other than the military and law enforcement personals. To construct the walkie-talkie system, the 2.4 GHz industrial, scientific, and medical (ISM) band and GFSK modulation technology have been chosen. In addition, the device includes transceiver module, dc power source, microphone amplifier, audio amplifier, dc-dc booster, and speaker. The coverage range of the implemented walkie-talkie system has been determined and found up to 200 meters. The gadget incorporates a li-ion battery charging system. The voltage, current and power levels at various locations of the device have been measured. It has been found that the maximum operating power consumption is 4.81W and standby power is 1mW. Moreover, the waveforms of different signals at different points of the systems have been observed on the oscilloscope. All of the development process of the system and the performance data are presented in this paper.

DUJASE Vol. 7 (2) 16-20, 2022 (July)

Improved Detection and Analysis of Wheat Spikes U sing Multi-Stage Convolutional Neural Network

M A Batin — 2023-04-04

High throughput plant phenotyping is the advanced scientific approach for rapid phenotyping of plant traits, especially high consumable grains or crops, which is designed to process a high volume of data in a short time for plant breeders and cultivars to utilize. Detection and counting of crop traits such as plants, fruits, wheat or rice spikes, sorghum head, and plant diseases is more advanced research in this field, where real-world data are collected using aerial and land-based imaging platforms equipped with a variety of geospatial sensors, and their statistical analysis is conducted using Artificial Intelligence (AI) and Deep Learning-based solutions. In this paper, we contributed to solving such a challenge of phenotyping by detecting and counting wheat spikes from land-based imaging by applying a Region-based Convolutional Neural Network (CNN) model. Our method employs the use of CNN to extract features from the imaging platform and the learning model is trained to detect and count wheat spikes in field images based on these extracted features. Using the publicly available SPIKE dataset to train and test our model, our proposed method achieved 98% average precision and 91% average F1 score on the test set. Our results show a significant improvement of 2.9% and 11.2% in detection accuracy as well as 1% and 3% in average precision metric over state-of-the-art Faster Regionbased Convolutional Neural Network (Faster-RCNN), and RetinaNet, respectively, and have the potential to significantly benefit plant breeders by facilitating the selection of wheat varieties with high yields.

DUJASE Vol. 7 (2) 21-30, 2022 (July)

Robust Signal Denoising Techniques for Highly Accurate and Spatial Resolution Preserved Brillouin Optical Time Domain Analysis Sensors

Abul Kalam Azad — 2023-04-04

The diminution of measurement uncertainty to an acceptable level and the preservation of experimental spatial resolution are highly desirable in many real-world applications of Brillouin optical time domain analysis (BOTDA) sensors. In practical applications of such sensors, the measurement uncertainty essentially relies upon the signal-to-noise ratio (SNR) of the experimental Brillouin gain spectra (BGSs) obtained throughout the sensing fiber. The improvement of such SNR using improper signal denoising techniques alters the experimental spatial resolution of the BOTDA sensors. In this paper, the use of non-local means filter (NLMF) and anisotropic diffusion filter (ADF) is experimentally demonstrated to enhance the uncertainty in the measurement of temperature using BOTDA sensors. For this purpose, the BGSs along a 41 km fiber are collected by averaging several numbers of BOTDA-traces from BOTDA hardware setup. Such BGSs are first denoised by employing NLMF and ADF to improve the measurement SNR. The Brillouin frequency shifts (BFSs) of denoised BGSs are then extracted via curve fitting technique (CFT). The BFS distribution is finally mapped to temperature distribution depending on the known BFS-temperature relationship of the sensing fiber. The robustness of the used filters is analyzed rigorously in terms of SNRs of BGSs, uncertainty in temperature extraction, spatial resolution and runtime in signal processing. The results indicate that the utilization of NLMF and ADF can enhance the SNRs of BGSs up to the maximum of 15.64 dB and 13.53 dB, respectively. Consequently, the uncertainties in the temperature extraction can be reduced up to the maximum of 52.63% and 57.31% for using NLMF and ADF, respectively. Moreover, both NLMF and ADF can preserve the experimental spatial resolution of the BOTDA sensors and include insignificant runtime to CFT. Thus, NLMF and ADF can be considered as robust signal denoising techniques for highly accurate and spatial resolution preserved BOTDA sensors.

DUJASE Vol. 7 (2) 31-38, 2022 (July)

Elucidation of Effectiveness of Dye Absorption as a Probe to Determine The Chemical Reactivity of Cellulose Nanofibrils

Khandoker Samaher Salem — 2023-04-04

Cellulose nanofibrils (CNF) has earned substantial attention as a sustainable biopolymer due to environmental consciousness and the implementation of strict regulations for the remediation of single use plastics. However, the existing processing strategies rely on intensive mechanical grinding and quality of production is often measured by physical changes such as fines which completely overlooks chemical changes such as exposure of large numbers of hydroxyls which significantly compromises desirable properties. Therefore, we use dye absorption technique to measure the chemical reactivity of the CNF. The CNF was produced from bleached hardwood kraft pulp (BHK) with different extent of fibrillation by varying the cumulative mechanical energy and then acetylation and dye absorption were used as a tool to investigate its chemical reactivity. The degree of substitution (DS) of the CNF reached a highest value and subsequently, it decreased to lesser values though the CNF were produced using higher cumulative energies. The dye absorption by CNF samples also followed the similar trend like the DS value of the CNF samples. The dye absorption was affected by the hydration shell formed by the hard-to-remove (HR) water molecules and the self-aggregation of cellulose hydroxyls at higher fibrillation, which hindered accessibility of the dye molecules to absorption cites. Therefore, this technique could be used as a new useful probing tool to determine the nanocellulose reactivity and might be a potential physicochemical strategy to produce and chemically modify CNF in an efficient and sustainable way to optimize its performance for different applications.

DUJASE Vol. 7 (2) 39-46, 2022 (July)

A Clustering based Feature Selection Approach using Maximum Spanning Tree

Md Hasan Tarek — 2023-04-04

Mutual information (MI) based feature selection methods are getting popular as its ability to capture the nonlinear and linear relationship among random variables and thus it performs better in different fields of machine learning. Traditional MI based feature selection algorithms use different techniques to find out the joint performance of features and select the relevant features among them. However, to do this, in many cases, they might incorporate redundant features. To solve these issues, we propose a feature selection method, namely Clustering based Feature Selection (CbFS), to cluster the features in such a way so that redundant and complementary features are grouped in the same cluster. Then, a subset of representative features is selected from each cluster. Experimental results of CbFS and four state-of-the-art methods are reported to measure the excellency of CbFS over twenty benchmark UCI datasets and three renowned network intrusion datasets. It shows that CbFS performs better than the comparative methods in terms of accuracy and performs better in identifying attack or normal instances in security datasets.

DUJASE Vol. 7 (2) 47-55, 2022 (July)

Design and simulation of Zeonex Based Suspended Microstructure Photonic Crystal Fiber for Chemical Sensing Application

Md Rakibul Islam — 2023-04-04

Over the last few years for sensing applications in terahertz regime photonic crystal fiber (PCF) has gained attention quite extensively. The optical characteristics of photonic crystal fiber can be controlled by fine-tuning of the structural parameters like core radius and effecctive area. In this context, a terahertz sensor based on a hollow-core photonic crystal fiber has been developed for chemical identification in the terahertz frequency range with very low loss. The proposed structure contains hexagonal manner sectored suspension type cladding and hexagonal manner sectored core region, all the sector are formed by zeonex based struts. To investigate the optical characteristics of developed design, finite element method (FEM) based COMSOL multiphysics v.5.4a software has been used. The simulation result shows the sensitivity of 83.37% and 83.63% at the optimum condition in x-polarization mode for ethanol and benzene respectively with low effective material loss of 0.0291 cm−1 and low confinement loss of 1.87×10-13 cm-1. Moreover, the developed design implementation is possible in the existing fabrication method. Physical features and comparative performance analysis are also showed in this research.

DUJASE Vol. 7 (2) 56-61, 2022 (July)

InGaN Laser Diodes-Gain, Spectra and Nonlinear Dynamics

Nahid Hassan — 2023-04-04

In this era, short-wavelength laser diodes with quantum-well (QW) structures offer plenty of opportunities for improvement in laser performance and receive widespread attention. Therefore, we develop dynamic model of the violet-blue InGaN laser diodes (LDs) and discuss basic features of the device in this paper. We investigate longitudinal mode dynamics through detail numerical simulation of the rate equation model of the quantum-well LDs. We selectively observe several effects such as relaxation oscillation, mode competition, intensity noise etc. From the dynamic behavior of the gain spectrum and time-varying modal photon numbers we found that the higher intensity noise of the quantum-well structure was due to random fluctuation with time among the different modes. The results were explained considering the previously published findings to confirm the validity of the proposed rate equation model of the quantum-well structures.

DUJASE Vol. 7 (2) 62-67, 2022 (July)